Abstract: A high-definition map system receives sensor data from vehicles traveling along routes and combines the data to generate a high definition map for use in driving vehicles, for example, for guiding autonomous vehicles. A pose graph is built from the collected data, each pose representing location and orientation of a vehicle. The pose graph is optimized to minimize constraints between poses. Points associated with surface are assigned a confidence measure determined using a measure of hardness/softness of the surface. A machine-learning-based result filter detects bad alignment results and prevents them from being entered in the subsequent global pose optimization. The alignment framework is parallelizable for execution using a parallel/distributed architecture. Alignment hot spots are detected for further verification and improvement. The system supports incremental updates, thereby allowing refinements of subgraphs for incrementally improving the high-definition map for keeping it up to date.

Abstract: A high-definition map system receives sensor data from vehicles traveling along routes and combines the data to generate a high definition map for use in driving vehicles, for example, for guiding autonomous vehicles. A pose graph is built from the collected data, each pose representing location and orientation of a vehicle. The pose graph is optimized to minimize constraints between poses. Points associated with surface are assigned a confidence measure determined using a measure of hardness/softness of the surface. A machine-learning-based result filter detects bad alignment results and prevents them from being entered in the subsequent global pose optimization. The alignment framework is parallelizable for execution using a parallel/distributed architecture. Alignment hot spots are detected for further verification and improvement.

Abstract: An exemplary method for navigating a virtual camera in a three dimensional environment includes determining a point at an off-center location of a viewport of the virtual camera. The point corresponds to a user selection, and the virtual camera defines a field of view to view content of the three dimensional environment. The method also includes rotating the virtual camera toward the off-center location such that the point is located at a center location of the viewport. The method further includes zooming the virtual camera toward the point located at the center location of the viewport. The method also includes after the zooming the virtual camera, re-rotating the virtual camera such that the point is located at the off-center location of the viewport. The method further includes after the re-rotating, rendering the content from a projection of the viewport.